Highly Size-Controlled, Low-Size-Dispersity Nickel Nanoparticles from

Dec 7, 2010 - E-mail: [email protected]. .... or alumina: a molecularly dispersed nickel precursor for syntheses of supported small size nickel nanoparti...
2 downloads 0 Views 4MB Size
Chem. Mater. 2010, 22, 6555–6563 6555 DOI:10.1021/cm102030p

Highly Size-Controlled, Low-Size-Dispersity Nickel Nanoparticles from Poly(propylene imine) Dendrimer-Ni(II) Complexes Elisabeta Mitran, Barry Dellinger, and Robin L. McCarley* Department of Chemistry, Louisiana State University, Baton Rouge, Louisiana 70803-1804, United States Received July 21, 2010. Revised Manuscript Received November 11, 2010

Reported here is the highly controlled synthesis and subsequent characterization of low-sizedispersity ((8-21%) crystalline Ni(0) nanoparticles derived from a dendrimer-ligand-based method employing amine-terminated poly(propylene imine) dendrimers, DAB-Amn. Crystalline Ni(0) nanoparticles devoid of any nickel boride are obtained by anaerobic borohydride reduction in methanol of DAB-Amn-Ni(II)x, as determined by high-resolution X-ray photoelectron spectroscopy (XPS), high-resolution transmission electron microscopy (HR-TEM), and selected-area electron diffraction (SAED) measurements. Nickel nanoparticles with highly tunable diameters ranging from l.9 to 2.7 nm are readily obtained by methanolic borohydride reduction of Ni(II) coordinated to the primary amines of five different generations of DAB-Amn (n = 4, 8, 16, 32, 64). Control over the diameter of the Ni nanoparticles is readily achievable and inversely related to dendrimer generation and the NH2:Ni(II) ratio, n/x. These outcomes bode well for future studies of relationships between metal nanoparticle properties and their behavior in the catalytic production of pollutants that have been found in combustion sources.

1. Introduction Particles produced in combustion chambers in the gas phase undergo inception, coagulation, and surface growth, and as a result, their size ranges from 2-4 nm for the smallest metallic nanoparticles, to a few hundred nanometers for agglomerates.1 Under primarily oxidative/oxidative pyrolysis of fuels, metal nuclei or supported metal nanoparticles may form and result in initiation of pollutant production on their surfaces.2 Several different transition metals are present in various fuels, including Cu, Fe, V, and Ni,1 and it is expected that the type and quantity of a given pollutant produced during the combustion process will be a function of metal identity and size. For example, Van den Brink et al.3 showed that in comparison to larger-sized Pt particles, supported Pt nanoparticles with very small size were more active in formation of polychlorinated benzene in the catalytic combustion of chlorobenzene. We propose that similar effects with Ni nanoparticles will occur, and there may exist relationships between particle size and both the nature and amount of pollutant products formed during particle-catalyzed reactions. However, what is needed is a simple route to the routine formation of zerovalent Ni nanoparticles in the 1-3 nm diameter range with tunable size control. *Corresponding author. Phone: (225) 578-3239. Fax: (225) 578-3458. E-mail: [email protected].

(1) Allouis, C.; Beretta, F.; D’Alessio, A. Chemosphere 2003, 51, 1091– 1096. (2) Walsh, M.; Cormier, S.; Varner, K.; Dellinger, B. EM Mag. 2010, April, 26-30. (3) van den Brink, R. W.; Krzan, M.; Feijen-Jeurissen, M. M. R.; Louw, R.; Mulder, P. Appl. Catal., B 2000, 24, 255–264. r 2010 American Chemical Society

The synthesis of metallic nickel nanoparticles is challenging because of their reactivity and the limitations of existing synthetic routes. Gas-phase routes include hydrogen reduction of NiCl2,4 hydrogen plasma metal reaction,5 and DC sputtering in argon,6 with the smallest nanoparticles having diameters d g 31 nm. Liquid-phase methods;such as microemulsion/surfactant techniques,7-10 chemical reduction in solution11-16 or in the presence of heterogeneous (4) Suh, Y. J.; Jang, H. D.; Chang, H. K.; Hwang, D. W.; Kim, H. C. Mater. Res. Bull. 2005, 40, 2100–2109. (5) Duan, H.; Lin, X.; Liu, G.; Xu, L.; Li, F. J. Mater. Process. Technol. 2008, 208, 494–498. (6) Rellinghaus, B.; Stappert, S.; Wassermann, E. F.; Sauer, H.; Spliethoff, B. Eur. Phys .J. D 2001, 16, 249–252. (7) Wu, S. H.; Chen, D. H. J. Colloid Interface Sci. 2003, 259, 282–286. (8) Chen, D. H.; Wu, S. H. Chem. Mater. 2000, 12, 1354–1360. (9) Legrand, J.; Taleb, A.; Gota, S.; Guittet, M. J.; Petit, C. Langmuir 2002, 18, 4131–4137. (10) Chiang, S.-J.; Liaw, B.-J.; Chen, Y.-Z. Appl. Catal., A 2007, 319, 144–152. (11) Hou, Y.; Kondoh, H.; Ohta, T.; Gao, S. Appl. Surf. Sci. 2005, 241, 218–222. (12) Chen, L.; Chen, J. M.; Zhou, H. D.; Zhang, D. J.; Wan, H. Q. Mater. Sci. Eng., A 2007, 452, 262–266. (13) Couto, G. G.; Klein, J. J.; Schreiner, W. H.; Mosca, D. H.; de Oliveira, A. J. A.; Zarbin, A. J. G. J. Colloid Interface Sci. 2007, 311, 461–468. (14) Roy, A.; Srinivas, V.; Ram, S.; De Toro, J. A.; Goff, J. P. J. Appl. Phys. 2006, 100, 094307. (15) Sidhaye, D. S.; Bala, T.; Srinath, S.; Srikanth, H.; Poddar, P.; Sastry, M.; Prasad, B. L. V. J. Phys. Chem. C 2009, 113, 3426–3429. (16) Kudlash, A. N.; Vorobyova, S. A.; Lesnikovich, A. I. J. Phys. Chem. Solids 2008, 69, 1652–1656. (17) Boudjahem, A. G.; Monteverdi, S.; Mercy, M.; Bettahar, M. M. Langmuir 2004, 20, 208–213. (18) Estournes, C.; Lutz, T.; Happich, J.; Quaranta, T.; Wissler, P.; Guille, J. L. J. Magn. Magn. Mater. 1997, 173, 83–92. (19) Chatterjee, A.; Chakravorty, D. Appl. Phys. Lett. 1992, 60, 138– 140.

Published on Web 12/07/2010

pubs.acs.org/cm

6556

Chem. Mater., Vol. 22, No. 24, 2010

supports,17,18 and organometallic routes,19 including thermolytic decomposition20-22;typically result in nanoparticles with d g 3 nm, and with size control only for particles above 5 nm.20,22,23 At present, we are unaware of a simple solution method, without using unimolecular templates such as dendrimers,24-30 that allows for careful control of Ni particle size over a range of sizes with d < 3 nm and with low size dispersity. Also, formation of Ni(0) particles free of boride contaminants, when using borohydride reducing agents in a variety of solvents, is a significant issue.9 To address these challenges, we turned to a nascent alternative methodology having the potential to form Ni nanoparticles with d < 3 nm, namely that which employs dendrimers.24 In an initial report, Knecht et al. prepared two sizes of ferromagnetic nickel nanoparticles with diameters of 0.8 ( 0.2 and 1.2 ( 0.3 nm by sodium triethylborohydride reduction of a 50% dodecyl-functionalized, sixth-generation PAMAM dendrimer coordinated to Ni(II) with toluene as solvent.24 We describe here a chemically distinct and simple synthetic route that leads to zerovalent, contaminant-free crystalline nickel nanoparticles, Ni(0)NP, whose diameters can be systematically tuned from 1.9 to 2.7 nm with a very low size dispersity. These unprecedented results for the formation of Ni(0) nanoparticles are achieved by use of a readily available, simple, and powerful hydride-ester-based reducing agent31 and unmodified, commercially available poly(propylene imine) dendrimers of five different generations. Specifically, this is accomplished by anerobic, methanolic borohydride reduction of stoichiometrically well-defined Ni(II)-dendrimer complexes that are based on coordination of Ni(II) to the dipropyltriamine groups at the periphery of the poly(propylene imine) dendrimers. By systematic control of the Ni(II)-dipropyltriamine ratio for a given dendrimer generation, as a function of the five different dendrimer generations, we are able to prepare highly tunable, statistically different-sized Ni nanoparticles with low size dispersity (10 min, the color of the solutions slowly dissipated, and this is the result of the oxidation of Ni(0) to (41) Scott, R. W. J.; Ye, H. C.; Henriquez, R. R.; Crooks, R. M. Chem. Mater. 2003, 15, 3873–3878.

Article

Figure 4. Absorption spectra of DAB-Am32-Ni(II)8 (A) before and (B) after reduction with NaBH4 in methanol under nitrogen environment, and (C) subsequent exposure to laboratory ambient. [Ni(II)] = 5  10-4 M.

Ni(II).24,42 The oxidation of Ni(0) to NiO is evidenced by loss of overall absorbance and the appearance of a new and characteristic absorption peak at 370 nm (Figure 4C) that is associated with NiO.43 DAB-Amn-Ni(0)NP from DAB-Amn-Ni(II)x via NaBH4 Reduction: Chemical Composition. XPS analyses of reduced DAB-Amn-Ni(II)x led to more detailed information about the oxidation state of Ni and the possible presence of other elements in the nanoparticles, such as nickel borides that are often observed in borohydrideproduced nanoparticles.9 Both survey (see Figure S1 in the Supporting Information) and high-resolution spectra for the Ni 2p and B1s regions were acquired. For comparison purposes, commercially available Ni “nanopowder” (diameter e 100 nm) and Ni2B powder (30 mesh) references were evaluated. Shown in Figure 5 is a representative high-resolution Ni 2p region spectrum of a typical borohydride-reduced DAB-Am16-Ni(II) sample supported on cleaned Pt foil (see Figure S2 in the Supporting Information). The metallic nature of the Ni atoms in the reduced sample is suggested by the 852.2 eV binding energy of the Ni 2p3/2 band; however, further examination of the Ni 2p1/2 and 2p3/2 and plasmon/shakeup loss satellite transitions44 allow for more definitive assignment of nickel valency. Confirmation of the zero-valence state of the nickel in the nanoparticles comes from the observed 17.2 eV binding energy difference between the Ni 2p1/2 and 2p3/2 transitions45 (Figure 5) and the 6.0 eV difference in the (42) Glavee, G. N.; Klabunde, K. J.; Sorensen, C. M.; Hadjipanayis, G. C. Langmuir 1994, 10, 4726–4730. (43) Qi, Y.; Qi, H.; Li, J.; Lu, C. J. Cryst. Growth 2008, 310, 4221–4225. (44) Grosvenor, A. P.; Biesinger, M. C.; Smart, R. S. C.; McIntyre, N. S. Surf. Sci. 2006, 600, 1771–1779. (45) Wagner, C. D.; Riggs, W. M.; Davis, L. E.; Moulder, J. F.; Muilenberg, G. E. Handbook of X-ray Photoelectron Spectroscopy; Perkin-Elmer Corporation Physical Electronics Division, 1979.

Chem. Mater., Vol. 22, No. 24, 2010

6559

Figure 5. Ni 2p high-resolution X-ray photoelectron spectra. Bottom: Nickel nanoparticles obtained by reduction of DAB-Am16-Ni(II)8 (x = n/2; [Ni(II)] = 5  10-3 M). Middle: Nickel nanopowder reference (d e 100 nm). Top: Ni2B reference.

Figure 6. B 1s high-resolution X-ray photoelectron spectra. Bottom: Nickel nanoparticles obtained by borohydride reduction of DABAm16-Ni(II)8 (x = n/2; [Ni(II)] = 5  10-3 M). Top: Ni2B reference.

2p3/2 and plasmon/shakeup loss satellite transitions (see Figure 5 and peak-fitted spectrum in Figure S3 in the Supporting Information), with the latter being extremely sensitive to the oxidation state and environment of the nickel.44 The difference in energy between the Ni 2p1/2 and 2p3/2 transitions in NiO has been observed to be 18.4 eV, whereas that in zerovalent nickel species is 17.4 eV, with the latter in good agreement with the 17.2 eV value observed for the DAB-Am16-Ni(0)NP.45 Importantly, the 6.0 eV difference in the 2p3/2 and plasmon/shakeup loss satellite transitions (852.2 and 858.2 eV) found for the DAB-Am16-Ni(0)NP is in agreement that observed for clean Ni(0) metal, 6.0 eV;44 NiO has its 2p3/2 transition

6560

Chem. Mater., Vol. 22, No. 24, 2010

Mitran et al.

Figure 7. Representative HR-TEM images and corresponding particle size distributions of Ni(0)NP formed by NaBH4 reduction of DAB-Am32-Ni(II)8, in MeOH: (A-C) for [Ni(II)] = 5  10-4 M and (D-F) for [Ni(II)] = 2  10-3 M. Insets are selected-area electron diffractograms.

centered at 854 eV and its plasmon/shakeup loss satellite located at 861 eV, whereas Ni(OH)2 exhibits values of 855.4 and 861 eV.44 Thus, we conclude that the Ni oxidation state is zero in the reduced DAB-Amn-Ni(II)x materials. We now turn our discussion to the well-known issue regarding the possible formation of metal borides during the borohydride reduction of metal salts, which has been observed when water or diglyme are used as solvents.9,42 In Figure 6 are presented representative high-resolution X-ray photoelectron spectra of DAB-Am16-Ni(0)NP and the Ni2B reference in the B 1s region. As expected for the Ni2B reference, a B-Ni band at 187.8 eV and that for B-O at ∼191.7 eV (in the boron oxide state) are present, and these values are in agreement with those of Ni-B alloys obtained by others.9,46-50 The presence of oxidized boron in the Ni2B reference is due to air oxidation, a wellknown phenomenon.9 Interestingly and importantly, the spectrum for DAB-Am16-Ni(0)NP does not possess the characteristic B-Ni band at 187.8 eV and thus does not support the presence of any boride (B-Ni) in the Ni nanoparticles. As a result, we conclude that the borohydride reduction of DAB-Amn-Ni(II)x, under the condi(46) Liu, Y. C.; Huang, C. Y.; Chen, Y. W. J. Nanopart. Res. 2006, 8, 223–234. (47) Okamoto, Y.; Nitta, Y.; Imanaka, T.; Teranishi, S. J. Chem. Soc., Faraday Trans. 1 1979, 75, 2027–2039. (48) Li, H.; Li, H. X.; Dai, W. L.; Wang, W. J.; Fang, Z. G.; Deng, J. F. Appl. Surf. Sci. 1999, 152, 25–34. (49) Lee, S. P.; Chen, Y. W. Ind. Eng. Chem. Res. 2001, 40, 1495–1499. (50) Liaw, B.-J.; Chiang, S.-J.; Tsai, C.-H.; Chen, Y.-Z. Appl. Catal., A 2005, 284, 239–246.

tions used here, results in zerovalent nickel nanoparticles free of metal boride. We postulate that this is the result of the methanol solvent employed, as it has been shown that borohydride reduction of Cu(II) in methanol results in Cu(0) particles, and not the boride.51 It should be noted that the spectrum for DAB-Am16Ni(0)NP does exhibit the characteristic B-O transition centered at 192 eV.9 This is not surprising, as no efforts were taken to remove the other products from the synthesis because of possible oxidation and loss of the metal nanoparticles. Because of the reaction of borohydride with protic methanol to yield either methylboronate or hydrated sodium borate and hydrogen gas,52 boron-oxygen species will always be present in the borohydride chemical reduction medium. Thus, it can be concluded that the peak at 192 eV for DAB-Am16Ni(0)NP is most likely associated with the presence of H3BO39,45,53 as a ∼10% impurity, but it is evidently not intimately associated with the Ni(0) nanoparticles based on results from electron microscopy studies, see below. DAB-Amn-Ni(0)NP from DAB-Amn-Ni(II)x via NaBH4 Reduction: Microscopically Determined Sizes. HR-TEM investigation of DAB-Amn-Ni(II)x (x = n/4) solutions treated with a 10 mol excess of borohydride routinely lead to images such as those in A and D in Figure 7. HR-TEM images reveal the crystalline structure of the Ni(0) nanoparticles (51) Jackelen, A.-M. L.; Jungbauer, M.; Glavee, G. N. Langmuir 1999, 15, 2322–2326. (52) Lo, C.-t. F.; Karan, K.; Davis, B. R. Ind. Eng. Chem. Res. 2007, 46, 5478–5484. (53) Il’inchik, E. A.; Volkov, V. V.; Mazalov, L. N. J. Struct. Chem. (Transl. Zh. Strukt. Khim.) 2005, 46, 523–534.

Article

Chem. Mater., Vol. 22, No. 24, 2010

6561

Table 1. Properties of Ni Nanoparticles as a Function of Dendrimer Generation upon NaBH4 Reduction of Methanolic DAB-Amn-Ni(II)x (4:1 = n/x) with Fixed [Ni(II)] = 5  10-4 M (BH4-:Ni(II) = 10:1)a sample

average Ni(0)NP diameter ( 1 s {%} (nm)

no. of Ni atoms per Ni(0)NPb

no. of DAB-Amn-Ni(II)x required to make 1 Ni(0)NP [aggregate diameter (nm)]c

Γ ΝH2on Ni(0)NP(mol cm-2)d

DAB-Am4-Ni(0)NP DAB-Am8-Ni(0)NP DAB-Am16-Ni(0)NP DAB-Am32-Ni(0)NP DAB-Am64-Ni(0)NP

2.7 ( 0.4 {15%} 2.6 ( 0.3 {12%} 2.3 ( 0.4 {17%} 2.1 ( 0.3 {14%} 1.9 ( 0.4 {21%}

963 890 611 428 325

963 [19] 445 [24] 153 [21] 54 [17] 20 [14]

5.58  10-8 5.41  10-8 4.80  10-8 4.26  10-8 3.89  10-8

a Mean diameters (( one standard deviation {% size dispersity}) are significantly different beyond the 99.9% confidence level (as judged by application of the t-test to the average particle diameters of the next closest-sized particles). The number of particles used for calculating the average and standard deviation is 800, 670, 465, 395, and 355 for the 1st-5th generation dendrimer cases. b Calculated according to Leff et al. in J. Phys. Chem. 1995, 99, 7036-7041. c The aggregate diameter was computed using the method in footnote a and employed estimated DAB-Amn-Ni(II)x diameters of 1.15, 1.90, 2.30, 2.72, and 3.14 nm for the n = 4-64 materials. d Estimated using the surface area of all nanoparticles generated by reduction and the total amount of dendrimer in solution.

Table 2. Properties of Ni Nanoparticles as a Function of Ratio of Primary Amines to Ni(II) Ions and Ni(II) Concentration in the Dendrimer Precursora

NP Sample DAB-Am32-Ni(0)NP

ratio of NH2 to Ni(II) n/x

[Ni(II)] (M)

average Ni(0)NP diameter ( 1 s {%} (nm)

no. of Ni atoms per Ni(0)NPb

no. of DAB-Am32-Ni(II) required to make 1 Ni(0)NP [aggregate diameter (nm)]c

Γ NH2 on Ni(0)NP (mol cm-2)d

2:1 4:1

5  10-4 5  10-4 2  10-3 5  10-4

2.4 ( 0.2 {8%} 2.1 ( 0.3 {14%} 2.7 ( 0.4 {15%} 1.9 ( 0.4 {21%}

702 428 996 310

44 [16] 54 [17] 125 [22] 78 [20]

2.51  10-8 4.26  10-8 5.65  10-8 7.66  10-8

8:1 a

Mean diameters (( one standard deviation {% size dispersity}) are significantly different beyond the 99.9% confidence level (as judged by application of the t-test to the average particle diameters of the next closest-sized particles). The number of particles used for calculating the average and standard deviation is 270, 395, 457, and 500 for the 2:1, 4:1, and 8:1 ratio scenarios. The BH4-:Ni(II) = 10:1. b Calculated according to Leff et al. in J. Phys. Chem. 1995, 99, 7036-7041. c The aggregate diameter was computed using the method in footnote a and employed an estimated DABAm32-Ni(II)x diameter of 2.72 nm. d Estimated using the surface area of all nanoparticles generated by reduction and the total amount of dendrimer in solution.

(Figure 7B, E) with an interplanar spacing of 0.20 nm along the (111) phase. The selected-area electron diffraction, SAED, patterns (insets in Figure 7) confirm the presence of metallic Ni, as noted by the diffraction rings from the (111), (200), (220), and (222) planes of pure Ni(0) having a face-centered-cubic (fcc) structure. No other crystalline materials can be identified in these HR-TEM and SAED images. These observations are independent of concentration of the Ni(II) in the dendrimer precursor. TEM images from control experiments of DAB-Amn solutions treated with methanolic NaBH4 did not possess any particulates. It is clear that Ni(0) nanoparticles have formed, thereby corroborating our XPS results. We first examined the impact of dendrimer generation on Ni(0)NP size at fixed [Ni(II)] and x = n/4, as well as BH4-: Ni(II). HR-TEM images and particle size distributions for all dendrimer generations studied are presented in Figures S4 and S5 in the Supporting Information. The average diameters of the Ni nanoparticles as a function of dendrimer generation are compiled in Table 1, and they range between 2.7 and 1.9 nm, with notably low size dispersities. Upon statistical treatment of the Ni(0)NP diameter data, it was found that the differences in diameters are significant beyond the 99.9% confidence level (t-test). It is clear that the particle size decreases with increasing dendrimer generation. Esumi and Amis have observed a dendrimer-generation

trend in Au nanoparticle size with PAMAM dendrimers as templates.54-56 Investigations of particle size as a function of the NH2: Ni(II) ratio in the dendrimer precursor and nickel chloride concentration were carried out, Table 2. Upon statistical treatment of the Ni(0)NP diameter data, it was found that the differences in diameters are significant beyond the 99.9% confidence level (t-test). From the results in Table 2 and Figure S6 in the Supporting Information, it is revealed that the size of the DAB-Amn-Ni(0)NP is affected by variation of the NH2:Ni(II) ratio at fixed Ni(II) concentration. A significant decrease in Ni(0)NP size (∼21%, or 2.26-fold decrease in number of Ni atoms) was obtained by increasing the NH2:Ni(II) ratio in the dendrimer precursor from 2 to 8 at fixed Ni(II) concentration, (Table 2 and Figure S6 in the Supporting Information). In addition, there is a roughly 30% increase in particle size with an increase of the [Ni(II)] from 5  10-4 to 2  10-3 M (Figure 7). At this point, we turn to a discussion of Ni(0)NP formation from the DAB-Amn-Ni(II)x materials upon addition of BH4-. From the data in Tables 1 and 2, it is clear that the size dispersity of the Ni(0)NP is relatively invariant with respect to the various conditions for their preparation. Furthermore, the size dispersity is very small in all cases (8-21%). These results;according to nucleation theory57;point to the

(54) Esumi, K.; Suzuki, A.; Yamahira, A.; Torigoe, K. Langmuir 2000, 16, 2604–2608. (55) Esumi, K.; Suzuki, A.; Aihara, N.; Usui, K.; Torigoe, K. Langmuir 1998, 14, 3157–3159.

(56) Gr€ ohn, F.; Bauer, B. J.; Akpalu, Y. A.; Jackson, C. L.; Amis, E. J. Macromolecules 2000, 33, 6042–6050. (57) Finney, E. E.; Finke, R. G. J. Colloid Interface Sci. 2008, 317, 351– 374.

6562

Chem. Mater., Vol. 22, No. 24, 2010

Mitran et al.

Scheme 2. Proposed Path for Formation of DAB-Amn-Ni(0)NP

separation of the nucleation event from that of the particle growth event, either temporally or spatially. That is to say, the variation in size of the nanoparticles is small because the production of critical nuclei is very near completion before the onset of nuclei enlargement by addition of Ni(0) atoms. Typically, this is achieved by either kinetic (temporal) or steric (spatial) constraints, wherein either reducing agent is added extremely quickly to the metal ion solution or the metal ions are sequestered in a fixed volume through use of template molecules/assemblies during the reduction process.57 Importantly, under the conditions used here, the BH4- reducing agent was added at a relatively slow rate (0.5 mL in 30 s); thus the low size dispersities observed here are attributed to a scaffold effect imparted by the DAB-Amn environment. A strong interaction of the dpt groups of DAB-Amn with Ni(II);K = 2.45  107 for methyl-dpt of Scheme 1 (R = CH3-)58;and the resulting Ni(0) surface will result in a small Ni(0)NP size dispersity and Ni(0)NP size, as noted in a study of Pt and Pd nanoparticles produced from a select group of DAB-Amn (n = 8, 16, 32) at extremely high primary amine:metal ion ratios (200 e n/x e 800).28 We observe very small Ni(0)NP sizes that are highly dependent on dendrimer generation (Table 1) and NH2: Ni(II) values of 2 e n/x e 8 (Table 2). Ni(0)NP size is inversely proportional to dendrimer generation and NH2: Ni(II) ratio (n/x). Because of the sizes of the DAB-Amn in solution, the Ni(0)NP produced here cannot reside completely inside the dendrimers, and it is concluded that the nanoparticles are coated with the DAB-Amn (interdendrimer), yielding a nanocomposite material, similar to what we proposed in an initial study with Cu(II)-PPI complexes and that proposed by Gr€ ohn et alia for PAMAM-Au(III) materials.25,56 For example, in the case of the fifth-generation dendrimer in Table 1, the resulting Ni nanoparticle has a (58) Goldberg, D. E.; Fernelius, W. C. J. Phys. Chem. 1959, 63, 1328– 1330.

diameter of 1.9 nm that is only slightly smaller than that of DAB-Am64 (∼2.5 nm); these observations are in accord with those from our work with DAB-Amn-Cu(0)NP nanocomposites.25 In addition, it is clear that the number of Ni atoms contained in a single Ni(0)NP is more than that available in an individual precursor, DAB-Amn-Ni(II)x, which upon initial examination would seem to indicate that the growth process of the Ni(0) nuclei is not well separated from nuclei formation; however, this cannot be the case as the size dispersity of the Ni(0)NP is very low. Finally, the inverse relationship between Ni(0)NP size and n/x (NH2: Ni(II) ratio, Table 2) would at first glance lead to the conclusion that the developing nanoparticle is stabilized by the presence of more amine groups on its surface (see last column in Table 2), but this argument does not hold for the generation-dependent effects observed in Table 1. Similar attempts on our part to describe the mechanism of DABAmn-Cu(0)NP formation were met with similar frustration, but in that study we did not have detailed data regarding the effects of NH2:metal ion ratio on nanoparticle size.25 We propose that the observed inverse relationship between Ni(0)NP size and both dendrimer generation and NH2:Ni(II) ratio (n/x), and the larger-than-expected Ni(0)NP size, are due to the presence of DAB-Amn-Ni(II)x complex aggregates, whose size is dependent on the concentration and generation of DAB-Amn-Ni(II)x in the methanol solutions. As outlined in Scheme 2, we posit that upon {DAB-Amn-Ni(II)x}Y exposure to reducing agent, nanoparticles of a dimension corresponding to the total number of Ni atoms initially in the aggregate (n*Y/x) are formed, and the resulting Ni(0)NP are protected from further growth by their presence within the dendrimer aggregate. The aggregates of Ni(II)-dendrimer complexes form either as a result of the Ni(II) coordinating to preexisting aggregates of the DAB-Amn (path A in Scheme 2), or they form by aggregation of the DAB-Amn-Ni(II)x complexes (paths B and C in Scheme 2). Upon reduction of the Ni(II)-complex aggregates, the concentration of the

Article

Chem. Mater., Vol. 22, No. 24, 2010

Ni(0)atom species is above the supersaturation value (“critical value”), and nucleation occurs in the {DAB-Amn}Y aggregate to yield DAB-Amn-Ni(0)NP. This model is in accord with the larger-than-expected sizes of the Ni(0)NP found here and those in our previous study of Cu(0)NP,25 as well as the small size dispersity of the nanoparticles. In preliminary dynamic light scattering studies of DAB-Amn in methanol with added Cu(II), we find evidence for aggregation of DABAmn-Cu(II)x, and the aggregation behavior is a function of both the concentration of the dendrimer-metal ion complex and the NH2:Cu(II) ratio (n/x).59 At concentrations above roughly 1  10-3 M of dendrimer NH2 groups, aggregates ∼35 nm in diameter were observed for DAB-Am16-Cu(II)x and DAB-Am32-Cu(II)x; this value is most likely an overestimate of aggregate size, because of the interfering effects of light absorption by the sample.60 We conclude that aggregate formation would also occur for Ni(II), based on its similar chemical properties in comparison to Cu(II).5 At this time, we cannot discern between paths B and C in Scheme 2, but it is clear that aggregation of the DAB-Am32-Ni(II)x would lead to the results noted in Tables 1 and 2. Conclusions Ni(II) has been shown to form well-defined, stoichiometric complexes with DAB-Amn dendrimers in methanol. Using UV-vis spectroscopy and the method of continuous variation, as well as titrations, it was determined that the (59) Bantchev, G. B.; McCarley, R. L. 2010, unpublished results. (60) Schaertl, W.; Roos, C. Phys. Rev. E 1999, 60, 2020–2028.

6563

predominant complex between Ni(II) and all DAB-Amn dendrimer generations studied exhibits a 2:1 ratio of NH2 end groups to Ni(II). Crystalline Ni(0) nanoclusters possessing a fcc structure, with diameters ranging from 1.9 to 2.7 nm, have been prepared by methanolic NaBH4 reduction of Ni(II) coordinated to various generations of DAB-Amn dendrimers (1, 2, 3, 4, and 5) under a variety of conditions. From XPS characterization studies, it was found that the Ni nanoparticles prepared by the dendrimer-ligand-based method are in the metallic state without any Ni-B alloy present, and they possess an impurity of H3BO3 that forms during decomposition of NaBH4 in methanol solution. The size of the Ni(0) nanoparticles can be precisely controlled (in order of impact) via dendrimer generation or the ratio of primary amines to Ni(II) ions in the dendrimer precursor. We would like to expand the variety of metal ions that can be complexed by DAB-Amn dendrimers using the dendrimerligand-based method in order to create new hybrid materials, such as bimetallic nanoparticles supported on silica surfaces for environmental catalysis applications. Acknowledgment. The authors thank the National Science Foundation for financial support of this work and LSU Graduate School. In addition, we thank Dr. Dongmei Cao for assistance with obtaining TEM images. Supporting Information Available: XPS data, HR-TEM images, and particle size distribution of nickel nanoparticles (PDF). This material is available free of charge via the Internet at http://pubs.acs.org